Horizontal U-Tube
Heat Exchangers

Long enough for the saddle to slide the full calculated growth plus a margin, which for carbon steel is about 0.08 inch per 10 feet of saddle spacing for every 100°F above the installation temperature. Austenitic stainless shells grow about half again as much. The bolt in the slot is snugged rather than torqued, and often sits on a slip plate or lubricant so it moves on the first warm-up. A slot cut to the bolt diameter plus a little clearance, which is common on general fabrications, is the usual mistake; it binds, and the shell then pushes on the nozzles.

The saddle nearest the piping that cannot flex, which is usually the one at the front head where the tube-side nozzles connect. Fixing that end keeps the heaviest and stiffest connections stationary and lets the bonnet end of the shell grow away from them. On a U-tube exchanger both tube-side nozzles are on the front head, so the choice is simpler than on a two-head design. If the shell-side piping is the rigid one instead, fix the saddle nearest the shell nozzles and tell the piping designer which end moves.

Yes, a slight slope toward the condensate outlet so condensate drains instead of pooling along the bottom and submerging the lower tubes. Flooded tubes do no condensing, so the exchanger loses surface, and on a steam-to-water U-tube that shows as a hunting temperature control valve and water hammer at startup. The slope is small, but the drain nozzle must be at the true low point of the installed unit rather than the drawing. Note the slope on your Request a Quote so nozzle and saddle heights suit it.

Enough for the insulation jacket to pass between the saddle web and the shell without being crushed, and for the nozzle flanges to clear the jacket thickness on the piping side. Saddles usually carry a wear plate so the insulation stops at the plate edge. Where insulation is wrapped over the saddle, the sliding saddle cannot move and the slots do nothing. Give the insulation thickness with the inquiry and the saddle height and nozzle projections will be set to suit, which is far cheaper than shimming the unit in the field.

On the shell for the assembled unit and on the tubesheet for the bundle alone, because the two are lifted at different times and have different centers of gravity. The assembled lift must clear the nozzles and saddles; the bundle lift has to balance a cantilevered assembly whose center of gravity sits well toward the tubesheet end. A lug sized for the complete exchanger is not automatically safe for a bundle pull, since the load path is different. Lug positions appear on the outline drawing, and a davit can take over the bundle lift on units pulled routinely.

Yes, provided the bundle pull path runs out from under the rack rather than along it, and no rack column sits in that path. The front head should face the open side. Headroom under a rack is usually enough for a shell on low saddles, but a davit arm on the front head swings through an arc and needs that arc free. Check that the condensate drain still reaches the trap by gravity from a low mounting height. Settle these on the plot plan before the shell diameter is final.